Disaster Recovery: Why and How to Test Your Backup Plan

June 2, 2015

Server crashes can affect any business and can stem from such mundane causes as human error, power outage and hardware failure. With so much potential for data loss, having a backup recovery plan in place would seem like a no brainer for most organizations. This is particularly true given the plethora of cloud-based and other digital solutions available today, making backing up easier than ever before.

Still, not all backup recover plans are as robust as they need to be. For example, a local backup or off-site cloud storage solution may provide some relief in the form of simple recovery of files, folder, and email data, but may be insufficient to address total server loss. Because of this, it’s vital for companies to test their ability to restore data and resume operations before an actual crisis hits.

Potential Problems with Your Backup and Recovery Plan

Testing your backup and recovery plan validates that all critical data is being backed up. It’s easy to add new servers and forget to add them to the backup script. Storage media can have defects, corrupting data and making it impossible to restore.

Testing the plan will help identify how long your operations will be down while data is recovered. It will also help plan how to handle data that was lost, key to resuming operations—there may be as little as a few minutes of lost data if you have replication, but as much as a full day’s worth if you do nightly backups. You’ll need to have a manual way to recover those lost transactions.

It’s also important to realize that backup and recovery is about applications, not just data. Sometimes you’ll need to recover on an entirely new server, which can raise compatibility issues. Having the ability to restore data isn’t meaningful if the application that reads it is unable to run. Compatibility issues also come into play if you are restoring older data, perhaps for analysis purposes or legal reasons.

How to Test Your Backup and Recovery Plan

There are several ways to test your backup and recovery plan, ranging from reviewing a checklist of steps to a full-blown simulation. The full-blown simulation is the most thorough test, though the most difficult to pull off, as it needs to be done without impacting normal production activity.

While you may think that backing up and restoring data is a purely technical task, it’s important to involve the business to ensure the full impact is understood and addressed.

Testing the backup process should verify that backups are performed on all critical systems and that no servers or data were unintentionally excluded. Review a list of all servers and all key file systems to ensure that nothing was omitted.
Testing the recovery process should verify that the data for the desired day can be located and is readable. Tech staff should know how to properly shut down applications, if needed, before restoring data, and the steps to load data properly from the backup media. The technical steps to restart the applications should also be known, and the business should know the procedures for recreating lost data and resuming normal operations.

During the test, the time taken to complete key steps should be recorded as well as any problems encountered during file restoration or application restarts. These should be reviewed following the test to identify areas of improvement to speed and smooth recovery when the plan is followed for real.

Disaster Recovery

Disaster recovery requires more than just the ability to backup and restore data—you might need to bring up a separate data center in case of a true disaster—but being able to load the new servers with your data and applications is critical. Testing the backup and recovery process is a critical part of ensuring you can remain in business even when disaster strikes.

The All Covered Difference

All Covered can help your business develop a backup solution that provides comprehensive disaster recovery and business continuity even in the event of total server failure. Our Server Backup solution uses an easy-to-deploy dedicated local appliance that incorporates local backup, cloud backup, object-level Exchange restore, and offers a local and cloud server business continuity option. Data is compressed, duplicated and encrypted on the local appliance – then immediately replicated on a cloud server in All Covered’s secure data center. With this combination of local and cloud-based backup, any file, folder or other data deleted from the primary server can be quickly restored from either the local appliance or from the Cloud.

put a plan in place

Alex Collins
IT Services Consultant